should have a mirror to permit readings without parallax. If the instrument is mounted in a dusty room it should be protected by an outside case of steel or wood having a glass front, as the presence of dirt filtering into the interior of the galvanometer may seriously affect its calibration. Both the wall type and the portable indicator, especially the latter, must be sufficiently robust to withstand reasonablv severe usage.
Resistance of Indicating Instruments.—When operated at the highest safe working temperatures most base-metal couples develop a maximum electromotive force of less than 50 to 70 millivolts and the LeChatelier couple an electromotive force of about 16 millivolts. In order to measure such small electromotive forces accurately a very sensitive indicator or millivoltmeter is required. On the other hand the instrument must be robust and able to withstand rough handling to which it is necessarily more or less subjected. These conditions of mechanical robustness and of high accuracy as a pyrometer indicator are difficult to satisfy.
The deflection of the pointer of the millivoltmeter is approximately proportional to the current flowing through the coil so that if the resistance of the total thermoelectric circuit is very low, relatively large currents are obtained, with a resulting high torque or turning moment on the movable coil. Thus when the current is high the^ construction of the indicator may be made very robust. Relatively strong springs, against which the turning moment of the coil carrying the current is balanced, may be employed and less attention need be given to the friction of the pivots in their bearings. The torque may be made high without greatly increasing the resistance of the circuit by using a large number of turns of copper wire in the coil. Thus it is not especially difficult to construct a sufficiently sensitive millivoltmeter of low resistance. The objection, however, from the pyrometric standpoint, to such an instrument used as a simple galvanometer has led to the development of indicators of considerable resistance. These are made by placing a high resistance of zero temperature coefficient, such as manganin, in series with the coil, and by increasing the number of turns on the moving coil to compensate for the decrease in sensitivity arising from the increased resistance. The so-called swamping resistance of zero temperature coefficient may be so proportioned relative to the copper that, due account; being taken of the temperature coefficient of elasticity of the springs, the instrument as a whole possesses a negligible temperature coefficient. The use of high resistance greatly diminishes the magnitude of the current flowing through the coil arid thus decreases the torque producing deflection. Hence attention must be given to the elimination of bearing friction as far as possible. Thus the instrument is necessarily more delicate than an indicator of low resistance and of the same type. The advantage of an instrument having a high resistance is that changes in the line resistance of comparatively large magnitude do not greatly affect the readings, as shown by the following table, the resistances of the indicators being 300 and 10 ohms respectively and the normal line resistance 2 ohms.
TABLE 7.—ERROR DTJE TO CHANGE IN LINE RESISTANCE
Line resistance, ohms Per cent error in indicator reading Error in degrees at 1,000°C.
300 ohms 10 ohms 300 ohms 10 ohms
1 2 3 5 +0.33 ±0.00 -0.33 -0.66 + 9.1 ± 0.0 - 7.7 -14 3 +3.3 ±0.0 -3.3 -6.6 + 91 ± 0 - 77 -143 _